Abstract

The standard rule of single privative modification replaces privative modifiers by Boolean negation. This rule is valid, for sure, but also simplistic. If an individual a instantiates the privatively modified property (MF) then it is true that a instantiates the property of not being an F, but the rule fails to express the fact that the properties (MF) and F have something in common. We replace Boolean negation by property negation, enabling us to operate on contrary rather than contradictory properties. To this end, we apply our theory of intensional essentialism, which operates on properties (intensions) rather than their extensions. We argue that each property F is necessarily associated with an essence, which is the set of the so-called requisites of F that jointly define F. Privation deprives F of some but not all of its requisites, replacing them by their contradictories. We show that properties formed from iterated privatives, such as being an imaginary fake banknote, give rise to a trifurcation of cases between returning to the original root property or to a property contrary to it or being semantically undecidable for want of further information. In order to determine which of the three forks the bearers of particular instances of multiply modified properties land upon we must examine the requisites, both of unmodified and modified properties. Requisites underpin our presuppositional theory of positive predication. Whereas privation is about being deprived of certain properties, the assignment of requisites to properties makes positive predication possible, which is the predication of properties the bearers must have because they have a certain property formed by means of privation.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.1(1-24)
Journal of Applied Logic ••• (••••)••••••
Contents lists available at ScienceDirect
Journal of Applied Logic
www.elsevier.com/locate/jal
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Iterated privation and positive predication
Bjørn Jespersen a, Massimiliano Carrara b, Marie Duží a
aVSB-Technical University of Ostrava, Department of Computer Science, 17. listopadu 15, 708 33
Ostrava, Czech Republic
bFISPPA Department, Section of Philosophy, University of Padua, P.zza Capitaniato 3, 35139 Padova,
Italy
a r t i c l e i n f o a b s t r a c t
Article history:
Available online xxxx
Keywords:
Iterated modification
Privative modification
Property negation
Contraries
Requisite property
Intensional essentialism
Transparent Intensional Logic
TIL
The standard rule of single privative modification replaces privative modifiers by
Boolean negation. This rule is valid, for sure, but also simplistic. If an individual
ainstantiates the privatively modified property (M F ) then it is true that a
instantiates the property of not being an F, but the rule fails to express the
fact that the properties (M F ) and Fhave something in common. We replace
Boolean negation by property negation, enabling us to operate on contrary rather
than contradictory properties. To this end, we apply our theory of intensional
essentialism, which operates on properties (intensions) rather than their extensions.
We argue that each property Fis necessarily associated with an essence, which is
the set of the so-called requisites of Fthat jointly define F. Privation deprives Fof
some but not all of its requisites, replacing them by their contradictories. We show
that properties formed from iterated privatives, such as being an imaginary fake
banknote, give rise to a trifurcation of cases between returning to the original root
property or to a property contrary to it or being semantically undecidable for want
of further information. In order to determine which of the three forks the bearers
of particular instances of multiply modified properties land upon we must examine
the requisites, both of unmodified and modified properties. Requisites underpin our
presuppositional theory of positive predication. Whereas privation is about being
deprived of certain properties, the assignment of requisites to properties makes
positive predication possible, which is the predication of properties the bearers must
have because they have a certain property formed by means of privation.
© 2017 Elsevier B.V. All rights reserved.
1. Introduction
There are large amounts of natural-language text data that need to be analyzed and formalized, because
we want to build up question-answering systems over these data. We want not only to convey information
explicitly recorded in these texts but also to derive implicit information entailed by these explicit data so
as to answer questions in an intelligent way. In other words, we want to apply logical reasoning to these
E-mail addresses: bjorn.jespersen@gmail.com (B. Jespersen), massimiliano.carrara@unipd.it (M. Carrara), marie.duzi@vsb.cz
(M. Duží).
https://doi.org/10.1016/j.jal.2017.12.004
1570-8683/© 2017 Elsevier B.V. All rights reserved.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.2(1-24)
2B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
natural-language corpuses. To this end, we must analyze natural-language sentences in a fine-grained way.
Since adjectives that denote property modifiers are part and parcel of our everyday vernaculars as well
as artificial languages, we need to logically analyze property modifiers as well. Privation being the most
complicated kind of property modification, the goal of this paper is a fine-grained analysis of privation
accompanied by rules governing reasoning about sentences that contain such modifiers.
Privative modification is an operation that forms negated properties from properties. It is one among
three kinds of negation:
privation, which applies to properties
the complement function, which applies to sets
the Boolean not, which applies to propositions-in-extension, i.e. truth-values.
When propositions are identified with (or at least modelled as) sets, then the complement function
subsumes propositional negation as a special case. Nothing in this paper hinges on this. What matters is
the contrast between privation, which is property negation and therefore an operation on intensions, and
set-theoretic negation, which takes a set to its complement and is therefore an operation on extensions.
The standard theory of modifiers is Montague Grammar, which is a typed version of model-theoretic
intensional logic. This paper provides an extension of this framework such that it is now possible to analyze
a particular sort of properties (or predicates, in the formal mode) that would previously fall outside the
purview of the framework. The paper also offers reasons for revising one of the existing rules; however, the
extension we provide can be incorporated without revising anything. We are building upon the work of not
least Coulson and Fauconnier [3], Horn [13–15], Iwańska [16], Jespersen [17], Kamp [22], Montague [25],
Partee [27], Primiero and Jespersen [28], while the background theory is based on Duží et al. [6,8].
Montague Grammar comes with a well-entrenched logic for single privation. This framework states its
logic for the various modifiers in the form of elimination rules. The rule of single privation amounts to
replacing the privative modifier by Boolean negation:
ais a fake banknote
ais not a banknote single privation
This rule is valid, because all that is required for validity is that the property (here, banknote) modified
by the privative modifier (here, fake) not be predicated of a, and the conclusion achieves at least this much.
However, the above rule misses the internal link between the property of being a banknote and the property
of being a fake banknote. We will probe further into this point below, but the basic idea is that a fake
banknote is not just some object or other that fails to be a banknote, but rather it is an object that must
have a host of properties in common with banknotes. Though both forged banknotes and, say, weathercocks
and zebras are not banknotes, there is an intuitive sense in which forged banknotes are somehow ‘closer to’
banknotes than are weathercocks and zebras. The challenge before us is to define privation in such a way
that it is made explicit what banknotes and forged banknotes have, and must have, in common.
Another problem with the rule of single privation is that it fails to extend to iterated privation, as
Boolean negation can replace a privative modifier only once. Here are some examples of predicates that
express iterated privation:
‘is an imaginary fake banknote’
‘is a former heir apparent’
‘is a former fallen angel’
‘weighs almost half a kilo’
‘is anything but a false friend’
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.3(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 3
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
‘is a theory of non-antisymmetric mereology’
‘is an imaginary, burned fake banknote’
For instance, Horn in [13, pp. 296–308], [14,15] ponders the logic and rhetoric of double negatives, e.g.
as expressed by ‘not un-F’ (‘not unhappy’, ‘not impolite’, etc.).1Is a not impolite remark a polite remark,
perhaps even a very polite remark, as per litotes (cf. [14, pp. 86ff]); or a remark that is neither polite nor
impolite, ending up in the neutral mid-interval? For a further example, consider so-called superdollars, which
are not US dollars, but counterfeit 100-dollar bills manufactured in, e.g., North Korea that are materially
(though not conceptually) well-nigh indiscernible from their genuine originals.2This particular occurrence
of ‘super’ in ‘superdollar’ has a privative effect, so the predicate ‘is a fake superdollar’ expresses double
privation.3A fake superdollar is a fraudulent imitation of what is already a fraudulent imitation. If somebody
collects first-degree counterfeit banknotes then they want a superdollar, and not a fake superdollar, which
exemplifies second-degree forgery by being a fake fake.4We all know that faking a fake will not return us
to the genuine original; but how do we know that? There is also the opposite direction: although you start
out with a 100-dollar bill, successfully passing it off as a fake superdollar to a collector of forged banknotes
of any degree, your 100-dollar bill has not transmogrified into a fake superdollar, despite being accepted as
one. But how do we know that? The answer we will be pursuing is that we know that because we know the
meaning of the respective predicates.
But what to replace Boolean negation with in order to develop a logic of iterated privation? We suggest
replacing Boolean negation with property negation. First, property negation operates on properties, just as
property modifiers do, so the intensional character of modification is carried through to negation. Second,
property negation obeys a logic of contraries rather than contradictories, which provides the kind of rule
that privative modification requires.
Let us take a closer look at privation. There are two material sources of privation. One is resultative and
hence diachronic: individual aonce was an F, but is no longer an F.5A recaptured fugitive (cf. [11]) once
was a fugitive, but is no longer one. Finished meals, burnt (not just charred) pieces of meat, and obsolete
banknotes all exemplify resultative privation. Given the actual laws of nature, neither a finished meal, nor
a burnt piece of meat can again become a meal or a piece of meat, whereas an obsolete banknote might be
restored to its previous glory as a banknote should the social institutions so favour it. The other source is
achronic: adid not start out as an Fand might never become an F, although it is possible that amight in
fact become an F, as when the relevant social institutions decree that such-and-such fake banknotes shall
henceforth acquire the status of valid tender, thus turning them into banknotes. Only this latter property
is extraneous to the property of being a fake banknote.
There are two formal sources of privation: either by way of first-degree or higher-degree modification.
Either a privative modifier modifies a property that has already been modified by a privative modifier, as
when imaginary is applied to fake banknote. Or a privative modifier modifies another privative modifier,
and the resulting modifier is applied to a property, as when anything but is applied to false and the resulting
modifier, anything but false, is applied to friend.6(In this paper we shall consider only first-degree iterated
1See Horn [13, pp. 38–41] for a historical survey of various takes on contrariety and predicate term negation.
2Cf. http://en.wikipedia.org/wiki/Superdollar.
3It is not an open-and-shut issue whether some modifiers are absolutely privative while the rest are context-sensitive by being
privative only with respect to some argument properties. Fake might be an example of the former, though we are issuing no
guarantee. Examples of the latter would include Nordic gold, which is not gold (but an alloy); fides punica, which is not trust (but
treachery); a baker’s dozen, which is not a dozen (but thirteen); and Rocky Mountain oysters, which are not oysters. See also [16]
on context-sensitive privatives.
4We could shift both the real McCoys and the fakes one level up with collectors collecting second-degree fakes and being fooled
by third-degree fakes; and so on up.
5Other dynamic examples of ‘stages of loss in the privative process’ and ‘incomplete realizations of possible privational histories’
(Martin [24, p. 439, 441, resp.]) would include going bald, i.e. progressing (or perhaps regressing) toward being almost or entirely
without hair.
6Anything but is a privative intensifier, just like very is a subsective intensifier, as in very good.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.4(1-24)
4B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
privation in the interest of brevity.) What we just described is double privation, but the theory readily
generalizes to triple, quadruple (etc.) privation, as when imaginary is applied to burned fake banknote.
In the light of the fact that privative modifiers can be nested, one may wonder: could we avail ourselves
of a rule that would calculate, for an arbitrary string of privative modifiers of two or more, whether the
root property Fis true of a? Our research shows that no such rule is forthcoming. There can be no unique
rough-and-ready rule for iterated privation.
Iterated privation issues instead in a trifurcation of cases:
(i) ais an F
(ii) afails to be an F
(iii) it is semantically indeterminate whether ais an F
The fact that this trifurcation emerges reflects the nature of privation. The first of two general points
bearing on privation is the negative one that privation is about what something is not, or fails to be. It
is about one or more properties that an object is deprived of. In particular, no theory of privatives should
predict that fake banknotes are extracted from sets of banknotes: fake banknotes are not banknotes that
are fake.7
But the second point is the positive one that there is substantially more to privation than deprivation.
Let Fbe a property, Mpa modifier privative with respect to F, and [MpF] the privatively modified property
that results from applying the modifier to the root property. The intuition we wish to capture is that when
an object has the property [MpF] then the object is—in some sense yet to be made clear—‘closer’ to having
Fthan are many or most other objects that lack the privatively modified property [MpF].8By way of an
example, a fake banknote is ‘almost’ a banknote, definitely barred from being one, yet it has a greater overlap
in terms of properties with a banknote than have most other objects. Fake banknotes must share a host
of properties with banknotes; otherwise they could not be fake banknotes in the first place, but would be
merely, say, colourful slips of paper. For instance, a banknote must mention the issuing authority, a currency
and a denomination. Therefore, a fake banknote must also mention an issuing authority, a currency and a
denomination. If a fake banknote sports, for instance, the words ‘ECB’, ‘EURO’, ‘100’ and is printed on
cotton-based paper with the look and feel of garden-variety banknotes then it lends itself to several instances
of what we call positive predication. Positive predication predicates properties of an object which the object
must instantiate and which are not privatively modified.
Positive predication appears to be less complicated vis-à-vis achronic privation than diachronic privation.
A burnt piece of meat is ash (inorganic matter) and in this second state not at all close to being meat (organic
matter), whereas a fake banknote must be close to being a banknote. However, we are able to put forward
a theory of positive predication with regard to objects that exemplify privatively modified properties of
either kind, because we offer a presuppositional theory of privation. The theory is presuppositional because,
for an object to exemplify a privatively modified property, it is presupposed that the object should already
exemplify other properties. By way of illustration, the property of being a former smoker comes with the
presupposition that, as a matter of analytic necessity, whoever currently instantiates it previously, but no
longer, instantiated the property of being a smoker. Or if ahas the property of being a Vatican cardinal
then a has also the property of being fluent in Latin. Beyond the well-rehearsed example of former smokers,
the theory extends to not only social artefacts like positions in a hierarchy of institutional power, but also
7This is to say that we do not stretch the meaning of ‘is a banknote’ so as to include fake banknotes among the banknotes. Partee
[27] suggests using coercion to do just that, such that it becomes meaningful to inquire whether a banknote is a real banknote or
a fake banknote. Jespersen [17, p. 544, fn. 14] and Duží et al. [6, p. 400, fn. 52] argue against Partee’s suggestion.
8Coulson and Fauconnier [3] and Iwańska [16] also think of privation both as the elimination of some, but not all, properties
(or concepts, features, etc.) inhering in privatively modified properties, and as the ‘blending with’ or ‘introduction of’ additional
properties so as to form new, hybrid properties like stone lion or toy elephant.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.5(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 5
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
technical artefacts like tools and scientific frameworks like taxonomies.9The presuppositional theory enables
us to infer conclusions about, say, dead whales (a dead whale being a dead mammal), disassembled watches
(a disassembled watch not being a timekeeping device) and burnt pieces of meat (a burnt piece of meat
having been previously a piece of meat). Importantly, each and every property we countenance has a host
of other properties associated with it. Thus, each instance of [Mp[MpF]] also comes with a host of adjacent
properties that their respective bearers must also bear.
We call the adjacent properties requisites.10 Our thesis is that privation is the deprivation of some, but
not all, requisites. The surviving requisites, together with some added ones, form the basis of positive
predication. The above trifurcation arises because the root property Fwill be one of the requisites of
some multiply privatively modified properties, while non-Fwill be one of the requisites of other multiply
privatively modified properties, whereas neither F, nor non-Fis among the requisites of still other multiply
privatively modified properties. In the third and final case, as far as the semantics of such properties goes,
there is no semantic fact of the matter as to which side of the fence acomes down on. Extra-semantic,
empirical investigation must, in each individual case, determine which side a given individual comes down
on. To give a taste of the trifurcation, here is an example of each of its three horns.
If somebody is anything but a false friend then they are a friend (and that to a very high degree) (i).
If something weighs almost half a kilo then it weighs less than a kilo and, therefore, does not weigh a
kilo (ii).11
If someone is a former heir apparent then either they are now the incumbent monarch or they are no
longer being even considered for the throne (iii).
The requisites of a given property enable valid reasoning from assumptions about privatively modified
properties. Philosophically speaking, associating requisites with properties amounts, in the case of privation,
to laying down at least some of what goes into being a wooden horse, a burnt wooden horse, a burnt fake
wooden horse, a fake burnt wooden horse, etc. Achieving the latter, philosophical, objective comes with a
fair amount of idealization while still requiring substantial philosophical justification.12 In this paper we rest
content with setting out the formal features of the framework within which we discuss iterated privation
and positive predication. Just to be clear, while we will be arguing for a particular elimination rule for
privatives, we will not attempt to put forward any introduction rules for privatives in the vein of:
P1,...,Pn
ais an [MpF]
For particular instances of Piand [MpF], such a rule would make explicit what the conditions are for
being, e.g., a fake banknote, or a wooden horse, or a malfunctioning toothbrush.13 Philosophy of technology
would make a great leap forward if particular instances of Picould be spelt out with the rigour required
9For the relevant theory of presuppositions, see [7,9,8].
10 The notion of requisite was conceived by Tichý and introduced in [30, p. 408]. It has subsequently been turned into a theory of
intensional essentialism. See Jespersen and Materna [21], Duží et al. [6, Ch. 3].
11 We are making the fairly uncontroversial assumption that when something weighs almost half a kilo then it weighs no more
than that. We want to blot out the kind of scenario where something that weighs exactly a kilo weighs also 900 grams, almost half
a kilo, etc., in virtue of a simple argument of downward monotonicity that also validates the ‘countdown’ inference that if you have
five fingers on your hand then you also have four (three, ..., zero) fingers, which still does not entail that you have fifteen fingers.
12 See [2].
13 It is not a matter of course that malfunctioning is privative. It is on the causal-role theory of technical function (what cannot
hammer cannot be a hammer), whereas it is subsective on the proper-function theory (a malfunctioning hammer was still designed
to hammer as its proper function). See Jespersen and Carrara [20]. An interesting study on malfunctioning software has been
recently provided by Floridi, Fresco and Primiero [12]. The authors distinguish between two kinds of malfunctioning software,
namely in terms of ‘negative’ dysfunction and ‘positive’ misfunction. They argue that while dysfunction is the core property of
malfunctioning technical artefacts, an executed software token cannot dysfunction, because it will always work in accordance with
its design. Yet it can, and often will, misfunction, because the design does not completely live up to the intended specification.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.6(1-24)
6B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
for an introduction rule. But, although the notion of requisite would come in handy, this very ambitious
enterprise is beyond the compass of this paper.14
The fundamental distinction among modifiers is typically considered to be one between the subsectives
and the non-subsectives.15 The former group would consist of the pure subsectives (that are governed by the
upwardly monotonic rule of right subsectivity, which amounts to eliminating the modifier and predicating
the surviving property) and the intersectives (that are governed by the rule of right subsectivity and a rule
of left subsectivity).16
Here is a brief comparison in prose of the four standard types of modifiers, where an index is an index of
empirical evaluation, such as a possible world or a world/time pair.
Pure subsectives. At every index a skilful surgeon is a surgeon.
Intersectives. At every index a round peg is round and is a peg.
Privatives. At no index is a fake banknote a banknote.
Modals. At some indices an alleged assassin is an assassin, and at some other indices an alleged assassin
is not an assassin.
Montague [25, p. 211] seeks to provide a uniform theory of modifiers (strictly speaking, of adjectival
phrases). Each modifier, according to Montague, is a property-to-property mapping.17 We subscribe to
this uniform account of the corpus of modifiers. We depart, however, from Montague’s contention that
these functions are meaning-to-meaning functions (ibid.). The contention, of course, makes perfect in Mon-
tague’s intensional framework in which intensions (functions whose domain are the logical space of possible
worlds) count as meanings.18 In our framework, meaning-to-meaning functions would be hyperintension-to-
hyperintension functions. We have such functions, but we do not need them here. We do need hyperinten-
sions, however, when working with modifiers: we need hyperintensions (meanings) when defining a couple
of key notions that go into defining modifiers. In a word, we are using hyperintensions in order to operate
on intensions.
It is relevant to compare modal and iterated privative modification, for in neither case is only one
conclusion possible. The modals require extra-semantic, empirical inquiry to establish, for each particular
instance, which of two ways the facts happen to go. Only empirical inquiry can decide which allegations of
being an assassin are true and which ones are false. The iterated privatives require intra-semantic inquiry
to establish, for each particular instance, which of the three ways the meanings go. If we land on the third
fork, then we need to get out of the armchair and into the field to establish which way the facts happen to
go.
Privation is literally radical modification, because the root property is modified away. Subsective modi-
fication, by contrast, enriches the root property, whether the modifier be intersective (e.g. round ) or purely
subsective (e.g. skilful). A peg, say, is qualified as a round peg, or conversely, something round is qualified
as a round peg; and a surgeon as a skilful surgeon. A layer of modification is added on top of the existing
requisites of the root property. Privation goes in the opposite direction by purging the root property of some
of its requisites. This is the crucial step toward explaining why a fake banknote fails to be a banknote. One
property that drops out is that of being valid tender, which comes with requisites of its own. Yet privation
14 See Del Frate [5] for conceptual discussion of a catalogue of engineering conceptions of malfunction.
15 See, e.g., Makinson [23, pp. 64–65] on the distinction between qualifiers and proper modifiers.
16 See Jespersen [17] for two rules of left subsectivity.
17 A topic we will not be delving further into here is how to decide for a given token of a given adjective whether it denotes a
property or a modifier. See, however, Siegel [29], Kamp [22], Montague [25], Beesley [1]. Schematically put, Montague pairs all
adjectives off with modifiers, Beesley pairs all adjectives off with properties, and Kamp pairs some adjectives off with modifiers
and the rest with properties.
18 We are glossing over the facts that Montague did not fully commit to s(i.e. combined world/time pairs) as a stand-alone type
on an equal footing with e(i.e. ‘entity’), t(i.e. truth-value), etc., and that Montague’s empirical indices were combined world/time
pairs.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.7(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 7
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
not only detracts, but also adds something. One property that gets added is that of being a forgery (i.e. a
fraudulent imitation of something or other), which also comes with requisites of its own. The crucial step
toward the presuppositional theory required for positive predication is that privation adds new requisites to
the purged set of requisites. Moreover, some of these new requisites contradict some of the original purged
requisites. This explains why we can predicate several properties of fake banknotes that they must have.
If we did not assign requisites to properties, we would be left with an exceptionally minimalist logic of
iterated privation. First of all, the replacement of privatives by Boolean negation can occur only once, as
we announced at the outset. Here is why. The standard rule of single privation lays down what to do when
it is true that ahas property [MpF]. The rule fails to state what to do when the premise is the negation
that ahas property [MpF]. This inference, therefore, is invalid:
¬[[MpF]a]
¬¬[F a]
If, counterfactually, the rule for privation had specified logical equivalence between [[MpF]a] and ¬[F a]
then the above argument would have come out valid. However, the rule of privation does not specify
equivalence; rather it specifies that [[MpF]a] entails ¬[F a]. It is also intuitive enough that the above
inference must come out invalid. If it did not, all instances of double privation would land on the first fork.
Thus, a fake rhinestone diamond would emerge as a diamond. So not only would the inference fail to be
truth-preserving by over-generating instances of the first fork, it would also leave no room for the other two
forks.
Secondly, therefore, in the interest of setting up a logic of iterated privation, we suggest replacing Boolean
negation by property negation, denoted by ‘non. This replaces contradictories by contraries, which makes
for a sufficiently weakened form of negation. Applied to single privation, the result is:
ais a fake banknote
ais a non-banknote single privation
When ais a fake banknote at some index then ais sent to the complement set of the set of banknotes
at the same index, though not to just anywhere in the complement, but to its particular subset of fake
banknotes. The good news is that we can reiterate non so as to form the property non-non-banknote. The
bad news is that [non [non F ]] would be the final word on iterated privation in the absence of requisites.
The above trifurcation would remain, but it would be impossible to decide which particular fork a particular
instance of iterated privation landed on. A logic of iterated privation that amounted to replacing privatives
by non would grind to a halt after having established the general insight that pairs of privatives yield
contraries.
The thesis, then, that we are arguing for can be condensed thus. A logic of iterated privation that invokes
requisite properties of privatively modified properties enables positive predication and is in a position to
land particular instances of multiply privatively modified properties on the right fork.
The rest of the paper is organized as follows. Section 2 sets out the relevant portions of our formal
semantic theory. Section 3 compares the logic of subsectives against the logic of privatives, introduces
property negation, and offers case studies of each of the forks of the trifurcation.
2. Logical foundations
In this section, we set out the formal framework within which we raise and solve the problem of iterated
privation. The framework is a fragment of Tichý’s Transparent Intensional Logic (TIL). The relevant frag-
ment is more or less continuous with Montague’s intensional logic and its myriad extensions. However, TIL
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.8(1-24)
8B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
has added a theory of modal modifiers (see Primiero and Jespersen [28]) to the Montagovian corpus, and
a spelt-out logic, including an introduction rule, for intersective modifiers (see Jespersen [17]), as well as a
general rule of left subsectivity applying to privatives and modals, intersectives and pure subsectives (see
Jespersen [17]; Duží et al. [6, §4.4]), which extends to single privation. The present paper is the third and
final of a trilogy of papers on how to model various states (especially malfunction) of technical artefacts by
means of property modifiers. The two preceding papers are Jespersen and Carrara [19,20].
2.1. Key definitions of Transparent Intensional Logic
We need definitions of the following basic notions:
Simple type theory. We need this definition in order to define both intensional and extensional entities.
Properties of individuals are typed as functions from possible worlds to functions from times to sets of
individuals, where sets are identified with their characteristic functions. Property modifiers are typed
as property-to-property functions. (Modifier modifiers are typed as modifier-to-modifier functions.)
Constructions. We need this definition for the following reasons. Constructions are (fine-grained and
structured) meanings; we define four of the altogether six constructions that make up the full inductive
definition of constructions. Furthermore, the definition introduces the formalism of TIL, which is based
on λ-abstracts.
Requisite. The requisite relation Req is a relation-in-extension between two properties Rand P, such
that, necessarily, whatever is (in) the extension of Pmust, as a matter of analytic necessity, also be (in)
the extension of R, though not necessarily conversely. We say that Ris a requisite of P.
Essence. The essence of a property Pis the set of its requisites which together define P.
Property negation. Property negation, non, allows iteration and obeys a logic of contraries.
Note that our theory is based on what we call intensional essentialism. The analytically necessary relation
of being a requisite of Pand being an element of the essence of Pobtains between intensional entities such as
properties, and not between extensional entities (such as individuals) and intensional entities. Consequently,
we subscribe to individual anti-essentialism: no individual has any purely contingent property necessarily.
By ‘purely contingent property’ we mean a non-constant property that does not have what we call an
essential core; e.g., the property of having exactly as many inhabitants as Prague is necessarily exemplified
by Prague, whatever number of inhabitants Prague may happen to have.19
We define the essence of a property as a set of its requisites that jointly define the property. For instance,
the property of being a mammal is related by the requisite relation to the property of being a whale.
Thus, necessarily, if the individual ahappens to be a whale at a world/time index of evaluation then ais
also a mammal at this world/time. It is an open question (epistemologically and ontologically speaking)
whether ais a whale. Establishing whether it is one requires investigation a posteriori. On the other hand,
establishing whether amust be a mammal in case ahappens to be a whale is a priori, the requisite relation
being in-extension and as such independent of what is true at any particular state of affairs. Comparing the
essences of a root property and a modified property enables us to define subsective and privative modifiers
in a new way that is an extension of previous definitions.
19 See Duží et al. [6, §1.4.2.1] for a classification of empirical properties and (ibid.: 68) for the notion of essential core.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.9(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 9
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Definition 1 (Simple type theory). Let Bbe a base, where a base is a collection of pair-wise disjoint, non-empty
sets. Then:
i) Every member of Bis an elementary type of order 1 over B.
ii) Let α,β1, ..., βm(m > 0) be types of order 1 over B. Then the collection (α β1...βm) of all m-ary partial
mappings from β1×... ×βminto αis a functional type of order 1 over B.
iii) Nothing is a type of order 1 over B unless it so follows from (i) and (ii).
Notation. That an object Ois of type α, i.e. belongs to the type α, will be denoted ‘O:α.
Remark. For the purposes of natural-language analysis TIL uses the following so-called objectual base B
consisting of the following atomic types:
o: the set of truth-values T, F;
ι: the set of individuals (the universe of discourse);
τ: the set of real numbers (doubling as discrete times);
ω: the set of logically possible worlds (the logical space).
Definition 2 (Constructions).
(i) Variables x,y, ... are constructions that construct objects (elements of their respective ranges) depen-
dently on a valuation v; they v-construct.
(ii) Where Xis an object whatsoever (an extension, an intension or a construction), 0
Xis the construction
Trivialization.0
Xconstructs Xwithout any change in X.
(iii) Let X,Y1, ..., Ynbe constructions. Then Composition [X Y1...Ym] is the following construction. If X
v-constructs a function fof type (αβ1...βm), and Y1, ..., Ymv-construct entities B1, ..., Bmof types β1,
..., βm, respectively, then [X Y1...Ym]v-constructs the value (an entity, if any, of type α) of fon the
tuple-argument hB1, ..., Bmi. Otherwise [X Y1...Ym] does not v-construct anything and so is v-improper.
(iv) The Closure [λx1...xmY] is the following construction. Let x1, x2, ..., xmbe pair-wise distinct vari-
ables v-constructing entities of types β1, ..., βm, respectively, and Ya construction typed to v-construct
an α-entity. Then [λx1...xmY] is the construction Closure (or λ-Closure). It v-constructs the follow-
ing function f: (αβ1. . . βm). Let v(B1/x1, ..., Bm/xm) be a valuation identical with vat least up
to assigning objects B1:β1, ..., Bm:βmto variables x1, ..., xm. If Yis v(B1/x1, ..., Bm/xm)-improper
(see iii), then fis undefined on hB1, ..., Bmi. Otherwise the value of fon hB1, ..., Bmiis the α-entity
v(B1/x1, ..., Bm/xm)-constructed by Y.
(v) Nothing is a construction, unless it so follows from (i) through (iv).
Remark. That a variable x v-constructs entities of a type αwill be referred to as ‘ranging over α’, denoted
by ‘xvα. We model sets and relations by their characteristic functions. Thus, for instance, () is the
type of a set of individuals, while (oιι) is the type of binary relations-in-extension between individuals.
Empirical expressions denote empirical conditions that may or may not be satisfied at some world/time
pair of evaluation. We model these empirical conditions as possible-world-semantic (PWS) intensions. PWS
intensions are entities of type (βω): mappings from possible worlds to an arbitrary type β. The type β
is frequently the type of the chronology of α-objects, i.e., a mapping of type (ατ ). Thus α-intensions are
frequently functions of type (α(τω)), abbreviated as ‘ατ ω . Extensional entities are entities of the arbitrary
type αwhere α6= (βω) for any type β. Where wranges over ωand tover τ, the following logical form
essentially characterizes the logical syntax of empirical language: λwλt [...w...t...].
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.10(1-24)
10 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Examples of frequently used PWS intensions are:
propositions of type oτ ω
properties of individuals of type ()τ ω
binary relations-in-intension between individuals of type (oιι)τ ω
individual offices (or roles) of type ιτω
Logical objects like truth-functions and quantifiers are extensional: (conjunction), (disjunction),
(implication) are of type (ooo), and ¬(Boolean negation) of type (oo). Since TIL has no syncategorematic
symbols, all the symbols in the TIL formalism denote functions, including quantifiers. The quantifiers α,
αare type-theoretically polymorphic total functions, just as in Montague Grammar, of type (o()), for
an arbitrary type α, and are defined as follows.
Definition 3 (Quantifiers). The universal quantifier αis a function of type (o()) that takes a class A
of α-elements to Tif Acontains all elements of the type α, otherwise to F. The existential quantifier α
is a function of type (o()) that takes a class Aof α-elements to Tif Ais a non-empty class, otherwise
to F.
Notational conventions.
• ‘x . . .’ serves as a shorthand for ‘[0λx . . .]’; similarly for ‘y’: all variable-binding is λ-binding, and
universal (existential) quantification is presented by means of Trivialization.
Below all type indications will be provided outside the formulae in order not to clutter the notation.
The outermost brackets will be omitted whenever no confusion can arise.
While ‘X:α’ means that an object Xis (a member) of type α, ‘Xvα’ means that Xis typed to
v-construct an object of type α, if any. We write ‘Xα’ if no confusion concerning valuation arises.
wvωand tvτ.
If Cvατ ω then the frequently used Composition [[C w]t], which is the intensional descent (a.k.a.
extensionalization) of the α-intension v-constructed by C, will be encoded as ‘Cwt .
Predication is an instance of Composition.20 An empirical predicate such as ‘is a planet’ denotes the
property of being a planet; it is subsequently extensionalized in order to obtain the set of planets at the
empirical indices of evaluation; the characteristic function of the set is applied, by way of Composition, to
the individual of which the property of being a planet is predicated; the result (a truth-value) is finally
abstracted over by means of wand tvariables in order to construct an empirical truth-condition of type
oτ ω. The form of the predication of being a planet of an individual ais this:21
λwλt [0
P lanetwt 0
a]
The form of the predication of the subsectively modified property of being a gas planet is this:
λwλt [[0Gas 0
P lanet]wt 0
a]
To begin, construct, by way of Composition, the property of being a gas planet and then follow the same
steps as above.
20 See Duží et al. [6, §2.4.2] on predication.
21 We apply the method of analysis according to which semantically simple predicates like ‘is a planet’ are associated with the
Trivialization of the denoted object; 0P lanet, in this case. See Duží et al. [6, §2.1].
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.11(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 11
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Types: Planet : ()τ ω;a:ι;Gas : ((()τ ω)(()τ ω )); [0Gas 0
P lanet]()τ ω ; [[0Gas 0
P lanet]wt 0a]vo;
λwλt [[0
Gas 0
P lanet]wt 0
a]oτ ω: the proposition that ais a gas planet.
2.2. Requisites
The requisite relations Req are a family of relations-in-extension between two intensions, so they are of
the polymorphous type (o ατ ω βτ ω ), with the possibility that α=β.22 Infinitely many combinations of Req
are possible, but for our purpose we will need just this one:
Req : (o()τ ω ()τ ω )
Req is a relation between two properties of individuals, such that one is a requisite of the other.
TIL embraces partial functions.23 Partiality gives rise to the following complication. The requisite relation
obtains necessarily, i.e. for all worlds wand times t, and so the values at this or that hw,tiof particular
intensions are irrelevant. But the extensions of properties (i.e. sets) are isomorphic to characteristic functions,
and these functions are amenable to truth-value gaps. As already mentioned, the property of having stopped
smoking comes with a bulk of requisites including not least the property of being a former smoker. Thus,
the predication of such a property Pof amay also fail, causing [0
Pwt 0
a] to be v-improper. There is a
straightforward remedy, however, namely the propositional property of being true at hw,ti;True: (o oτ ω )τ ω .
Given a proposition Prop, [0T ruewt 0
P rop]v-constructs Tif Prop is true at hw,ti; otherwise (i.e., if Prop is
false or else undefined at hw,ti) it v-constructs F.
Definition 4 (Requisite relation between ι-properties). Let X,Ybe constructions such that X,Y(o ι)τ ω ;
xι. Then
[0
Req Y X] = wt[x[[0T ruew t λwλt [Xwt x]] [0T r uewt λwλt [Ywt x]]]].
Gloss definiendum as, “Yis a requisite of X”, and definiens as, “Necessarily, i.e. at every hw,ti, any x
that instantiates Xat hw,tialso instantiates Yat hw,ti.
Example. Let the property of being a person be a requisite of the property of being a student. Then the
hyperproposition that all students are persons is an analytic truth. It constructs the proposition TRUE,
which is the necessary proposition, which takes value Tat all world/time pairs. Wherever and whenever
somebody is a student they are also a person. Formally:
[0
Req 0
P erson 0
Student] = wt[x[[0T r uewt λwλt [0
Studentwt x]] [0T ruewt λwλt [0
P ersonwt x]]]]
Claim 1. Req is a quasi-order on the set of ι-properties.
Proof. Let X,Y()τ ω . Then Req belongs to the class QO:(o(o()τ ω ()τ ω)) of quasi-orders over the
set of individual properties:
22 For comparison, Jespersen [18] offers a detailed study of a requisite relation, of type (o ιτω ιτω ), where one individual office is a
requisite of another individual office, the way the office of Commander-in-Chief is a requisite of the office of President of the United
States. The paper analyses “Superman is Clark Kent” as expressing that this particular requisite relation obtains between one
office denoted by ‘Superman’ and another office denoted by ‘Clark Kent’. If you occupy the office of Superman you must co-occupy
the office of Clark Kent, but you can occupy the Clark Kent office without occupying the Superman office. This goes to show
that TIL offers an intensional analysis (based on intensional essentialism) of “Superman is Clark Kent”, contrary to the prevalent
‘Millian’ extensional analyses.
23 See Duží et al. [6, 276–78] for a philosophical justification of partiality in spite of the associated technical complications.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.12(1-24)
12 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Reflexivity. [0
Req X X ] = wt[x[[0T ruewt λwλt [Xwt x]] [0T ruewt λw λt [Xwt x]]]].
Transitivity. We want to prove that [[[0
Req Y X][0
Req Z Y ]] [0
Req Z X]].
1. [[0
Req Y X][0
Req Z Y ]]
2. [wt[x[[0T ruewt λwλt [Xwt x]] [0T ruewt λw λt [Ywt x]]]]
wt[x[[0T ruewt λwλt [Ywt x]] [0T ruewt λw λt [Zwt x]]]]] 1, Definition 4
3. [[0T ruewt λwλt [Xwt x]] [0T ruewt λwλt [Ywt x]]] 2, E,E
4. [[0T ruewt λwλt [Ywt x]] [0T ruewt λw λt [Zwt x]]] 2, E,E
5. [[0T ruewt λwλt [Xwt x]] [0T ruewt λwλt [Zwt x]]] 3, 4, *
6. [wt[x[[0T ruewt λwλt [Xwt x]] [0T ruewt λw λt [Zwt x]]]] 5, I
7. [[0
Req Z X] 6, Definition 4
8. [[[0
Req Y X][0
Req Z Y ]] [0
Req Z X]] 7, I
Remark. In line (5) ‘*’ denotes the theorem of the transitivity of implication.
In order for a requisite relation to be a weak partial order, it would need to be also anti-symmetric. The
Req relation is, however, not anti-symmetric. If properties X,Yare mutually in the Req relation, i.e. if
[[0
Req Y X][0
Req X Y ]]
then at every hw, tithe two properties are true of exactly the same individuals. This does not entail, however,
that X,Yare identical. It may be the case that there is an individual asuch that [Xwt a]v-constructs F
whereas [Ywt a] is v-improper. For instance, the following properties X,Ydiffer only in truth-value for those
individuals who never smoked. Let StopSmoke:()τω be the property of having stopped smoking. Whereas
Xyields truth-value gaps on such individuals, Yis false of them:
X=λwλt λx [0
StopSmokew t x]
Y=λwλt λx [0T ruewt λw λt [0
StopSmokew t x]]
This makes for a negligible difference that can be abstracted away, so we introduce the equiva-
lence relation : (o()τ ω()τ ω ) on the set of individual properties; p, q ()τ ω; = : (ooo); =df :
(o(o()τ ω()τ ω )(o()τ ω ()τ ω )), i.e. the identity of binary relations between properties.
0=λpq [x[[0T ruewt λwλt [pwt x]] = [0T ruewt λw λt [qwt x]]]]
Now we can define the Req’ relation on the factor set of the set of ι-properties as follows.24
Let [p]=λq [0p q] and [Req[p][q]] = [Req p q ]. Then:
Claim 2. Req’ is a partial order on the factor set of the set of ι-properties with respect to the relation .
Proof. It is sufficient to prove that Req’ is well-defined. Let p,qbe ι-properties such that [ 0p p] and
[0q q]. Then:
[Req[p][q]] = [Req p q]
=wt[x[[0T ruewt λwλt [pwt x]] [0T ruewt λw λt [qwt x]]]]
=wt[x[[0T ruewt λwλt [p
wt x]] [0T ruewt λwλt [q
wt x]]]] = [Req[p][q]]
24 The definition of Req’ was first introduced in Duží et al. [6, 363–364].
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.13(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 13
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Now, obviously, the relation Req’ is anti-symmetric:
[[0
Req[p][q]][0
Req[q][p]]] [[p]= [q]]
To make the exposition easier to follow, in what follows we will neglect this minor difference between
properties λwλt λx [0T r uewt λwλt [pwt x]] and pso that instead of the former we will write simply ‘p.
2.3. Intensional essentialism
Next, we are going to define the essence of a property. Our essentialism is based on the idea that since no
purely contingent property can be essential of any individual, essences are borne by intensions rather than
by individuals exemplifying intensions. That a property Fhas an essence means that a relation-in-extension
obtains a priori between Fand a set Essence of other properties such that, as a matter of analytic necessity,
whenever an individual (an ι-entity) instantiates Fat some hw , tithen the same individual also instantiates
all the properties belonging to Essence at the same hw, ti. Hence our essentialism is based on the requisite
relation, couching essentialism in terms of a priori interplay between properties, regardless of who or what
exemplifies a given property. The essence of a property Fis identical to the set of requisites that jointly
define F. The hw, ti-relative extensions of a given property are irrelevant, as we said; but so are the various
equivalent constructions of the property.
Definition 5 (Essence of a property). Let p, q ()τ ω ;Ess: ((o()τ ω )()τ ω ), i.e. a function assigning to a
given property pthe set of its requisites defined as follows:
0
Ess =λpλq [0
Req q p]
Then the essence of a property pis the set of its requisites:
[0
Ess p] = λq [0
Req q p]
Each property has many requisites. The question is: how do we know which properties are the requisites
of a given property? The answer requires an analytic definition of the given property. For instance, consider
the property of being a (domestic) cat. A classification according to biological taxonomy can serve as such
a definition:
Kingdom: Animalia
Phylum: Chordata
Clade: Synapsia
Class: Mammalia
Order: Carnivora
Family: Felidae
Subfamily: Felinae
Genus: Felis
Species: Felis Catus
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.14(1-24)
14 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Thus, we can define a cat as an animal belonging to all of the above categories.25 From this definition it
follows that, for instance, the sentence “Cats are mammals” comes out analytically true:
wt[x[[0
Catw t x][0
Mammalwt x]]]
Hence the property of being a mammal is a requisite of the property of being a cat. All the above properties
defined by a given taxonomy belong to the essence of the property of being a cat.
3. Subsectives, privatives, property negation, and case studies
3.1. Subsectives and privatives
With the above definitions in place, we can go on to compare two kinds of subsectives against privatives:26
A modifier Mis non-trivially subsective with respect to property Fiff the modified property [M F ] has
all the requisites of Fand at least one additional requisite that is not a requisite of F. In other words,
the essence of Fis a proper subset of the essence of [M F ].
For instance, a skilful surgeon is a surgeon because the property of being a skilful surgeon must have all
the requisites of the property of being a surgeon, and the additional property of being skilful with respect
to the property of being a surgeon.
A modifier Mis trivially subsective with respect to Fiff the modified property [M F ] has exactly the
same requisites as the property F, i.e. if [M F ] and Fshare the same essence, hence are identical
properties. The trivial subsectives are trivial in that the modification has no effect on the modified
property and so might just as well not have taken place.
For instance, there is no semantic or logical (but perhaps rhetorical) difference between the property of
being a diamond and the property of being a genuine diamond. Trivial modifiers such as genuine and real
are pure subsectives: genuine diamonds are not located in the intersection of diamonds and objects that are
genuine, for there is no such property as being genuine, pure and simple. Genuine diamonds form a subset,
though not a proper one, of a given set of diamonds.27
• A modifier Mis privative with respect to Fiff the modified property [M F ] lacks at least one, but
not all, of the requisites of the property F. Moreover, the essence of [M F ] contains at least one other
requisite that does not belong to the essence of F, and contradicts at least one of the requisites of F.
Hence, Mis privative with respect to Fiff the essence of [M F ] has a non-empty intersection with the
essence of F, and this intersection is a proper subset of both the essences of Fand of [M F ].
25 Contra Kripke, it is not a discovery (a posteriori, yet ‘metaphysically’ necessary) that a domestic cat belongs to any of the
categories above. The definition of domestic cat in virtue of the conjunction of the above categories is a stipulative definition,
which is conceptually prior to any empirical discovery of the further properties of various domestic cats (such as weighing seven
pounds, basking on a hot tin roof, or having grey stripes). Our stance is at odds with Kripkean essentialism, as we find anyone
conducting empirical inquiry in the animal kingdom needs a conceptual steer on what deserves to be called a domestic cat in the
first place before they can claim to have had any sort of causal interaction with domestic cats. (These remarks barely scrape the
surface of a deep philosophical issue, but they serve at least to indicate where we stand.)
26 We are disregarding intersective modification in order not to clutter the exposition. However, intersectives are controlled by the
same rule of right subsectivity that applies to the subsectives, together with the special rule of left subsectivity defined in [17].
27 Iwańska [16, p. 350] refers to ‘ideal’, ‘real’, ‘true’, and ‘perfect’ as type-reinforcing adjectives, which seems to get the pragmatics
right of what are semantically pleonastic adjectives. Trivial subsectives should not be confused with subsective intensifiers, as in
‘is real pain’, when real pain does not contrast with imaginary pain, but with slight pain.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.15(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 15
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
For instance, forged banknote has almost the same requisites as does banknote, but it has also another
requisite, namely the property of not being issued by an instance endowed with issuing authority.
To formally define the difference between subsective and privative modification, we need the TIL definition
of the relation of being a subset between sets and the operation of the intersection of two sets. The relation
of being a subset between α-sets, : (o()()), is defined for any type αas follows. Let a, b v(),
xvα. Then:
0
=λab [x[[a x][b x]]]
The relation of being a proper subset,: (o()()), is then defined as usual:
0
=λab [[x[[a x][b x]]] ∧ ¬[0
=a b]]
For instance, that the set of primes, Prime: ( ), is a subset of the naturals, Natural: ( ), is captured by
this construction:
[00
P rime 0
Natural] = [x[0
P rime x][0
N atural x]]
Similarly, that the set of primes is a proper subset of the naturals is captured by this construction:
[0
0
P rime 0
Natural] = [[x[0
P rime x][0
N atural x]] [0
P rime 6=0
Natural]]
The operation of intersection, : (()()()), is defined as follows:
0=λab λx [[ax][bx]]
For instance, that the intersection of primes and even numbers, Even: (), is equal to the singleton 2 is
captured by this construction:
[00
P rime 0
Even] = λx [[0
P rime x][0
Ev en x]] = λx [x=02]
In what follows we will use classical (infix) set-theoretical notation for any sets A,B; hence instead of
‘[0
A B]’ we will write ‘[AB]’, and instead of ‘[0A B]’ we will write ‘[AB]’. Since we will be comparing
sets of properties, the type αis here the type of an individual property, ()τω .
We are now able to provide the following two definitions.
Definition 6 (Subsective vs. privative modifiers). Let M(()τω ()τ ω ); F, p ()τω . Then
A modifier Mis subsective with respect to a property Fiff
[0
Ess F ][0
Ess [M F ]]
A modifier Mis non-trivially subsective with respect to a property Fiff
[0
Ess F ][0
Ess [M F ]]
A modifier Mis privative with respect to a property Fiff
[[0
Ess F ][0
Ess [M F ]]] 6=∅ ∧
p[[[0
Ess F ]p][[0
Ess [M F ]] λwλt [λx ¬[pwtx]]]]
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.16(1-24)
16 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Remark. The second conjunct defining privative modifier is to be read as follows: “There is a property
psuch that it is a requisite of the property F([[0
Ess F ]p]), and among the requisites of the modified
property [M F ] there is a property that contradicts p: [[0
Ess [M F ]] λwλt [λx ¬[pwtx]]].” This follows from
the semantics of privative modification. The privative modifier Mnot only deprives the property Fof one
or more of its requisites, it also adds at least one requisite that causes privation.
Remark. The above definition of subsective and privative modifiers is a novel contribution of this paper. It
is an improvement over the corresponding definitions in Primiero and Jespersen [28] and Duží et al. [6, §4.4].
As for subsective modifiers, the new definition differentiates between non-trivially and trivially subsective
modifiers. As for privatives, the original definition is a logical consequence of this new one, as we are going
to prove below. It not only stipulates that among the requisites of the privatively modified property Fis
the property of not being an F, but also explains why it is so. Furthermore, the new definition also specifies
what the modified property and the root property have in common. Privation deprives the root property
of some but not all of its requisites. The more requisites of the root property Fare preserved, the closer a
relative the modified property is to F. Thus, we are able to keep track of the root property in the modified
property, which in turn makes it possible to prove that, for instance, a demolished damaged house is not a
demolished damaged bridge (see below for this example).
Example. The modifier Wooden: (()τ ω ()τ ω ) is subsective with respect to the property of being a table,
Table: ()τ ω , but privative with respect to the property of being a horse, Horse: ()τ ω . Of course, a
wooden table is a table, but the essence of the property [0W ooden 0T able] is enriched by the property of
being wooden. Being wooden is a requisite of the property of being a wooden table, but it is not a requisite
of the property of being a table, because tables can be instead made of stone, iron, glass, etc.
[0
Ess 0
T able][0
Ess [0W ooden 0T abl e]]
But a wooden horse is not a horse. The modifier Wooden, the same modifier that just modified Table, deprives
the essence of Horse of many requisites, for instance of the property of being a living thing, or having a
bloodstream, or having kidneys, etc. Hence among the requisites of the property [0W ooden 0Horse] there
are properties like not being a living thing,not having a bloodstream, etc., which are contradictory (not just
contrary) to some of the requisites of the property Horse. On the other hand, the property [0W ooden 0H orse]
shares many requisites with the property of being a horse, like the outline of the body, resemblance of a
horse, etc., and has the additional requisite of being made of wood. Thus, we have (LT : ()τ ω , the property
of being a living thing, HB: ()τ ω , the property of having blood):
[[0
Ess 0
Horse][0
Ess [0
W ooden 0
Horse]]] 6=∅ ∧
[[[0
Ess 0
Horse]0
LT ][[0
Ess [0
W ooden 0
Horse]] λwλt λx ¬[0
LT x]]]
[[[0
Ess 0
Horse]0
HB][[0
Ess [0
W ooden 0
Horse]] λwλt λx ¬[0
HB x]]]
etc.
At the outset of this paper we characterized the difference between subsective and privative modifiers
by means of the rule of right subsectivity, which holds for subsective but not privative modifiers: a skilful
surgeon is a surgeon; a fake banknote fails to be a bank note.
When Ms(()τ ω ()τω ) is a construction of a modifier subsective with respect to the property
v-constructed by F()τω , then necessarily and for all individuals xthe following rule of right subsectivity
(RS) is valid:
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.17(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 17
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
[[MsF]wt x]
[Fwt x]RS
By Definition 6 it holds that [0Ess F ][0Ess [MsF]]. Hence each requisite of Fis also a requisite of [MsF],
but not vice versa, provided Msis non-trivially subsective. By Definition 4 and Claim 1, since each property
is a requisite of itself, it follows that Fis a requisite of [MsF]:
wt[x[[0T ruewt λwλt [[MsF]wt x]] [0T ruewt λwλt [Fwt x]]]]
which proves the rule of right subsectivity (RS).
For privatives, we already suggested replacing Boolean negation by property negation, denoted by ‘non’,
to specify the rule governing privatives. Let Mp(()τ ω ()τ ω ) be a construction of a modifier privative
with respect to the property v-constructed by F()τω . Then:
[[MpF]wt x]
[[non F ]wt x]Priv
Of course, it also holds that if xis an [MpF] then it is not the case that xis an F:
[[MpF]wt x]
¬[Fwt x]Single Privation
The reason for replacing Boolean negation by property negation is this. For each individual xand for
each property F, it is either true that xis an For it is not true. Yet there are many individuals that
are neither an Fnor a [non F ]. For instance, each individual either is or is not a banknote. Yet most
individuals are neither a banknote nor a fake banknote, because a fake banknote must still have something
in common with a banknote. A well-forged banknote is almost a banknote, because the property of being a
well-forged banknote is a ‘close relative’ of the property of being a banknote, sharing many requisites with
this property. Hence the property [non F ] is not contradictory but only contrary to F. Due to the difference
between contradictory and contrary properties, the Priv rule is indeterministic between the three forks with
the third fork having a further measure of indeterminacy, whereas the standard rule of single privation is
deterministic. Our strategy being that the non-based rule of privation ought to be extended to all instances
of single privation, the discrepancy between indeterministic and deterministic rules will vanish, as both the
rule of single and the rule of iterated privation will now be indeterministic.
We are now going to define the property negation non and prove that Priv is valid for privative modifiers.
3.2. Property negation
The philosophical source of inspiration is Aristotle’s observations that:
The sentences “It is a not-white log” and “It is not a white log” do not imply one another’s truth. For if
“It is a not-white log” is true, it must be a log: but that which is not a white log need not be a log at
all. (Prior Analytics I, 46, 1.)
That is, in modern parlance, a set of logs divides into those that are white and those that are non-white,
whereas a set of non-(white logs) divides into those elements that are non-white logs and those that are
not even logs (though perhaps white). More specifically, this quotation has inspired us to adopt property
negation. And directly relevant for our present purpose:
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.18(1-24)
18 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
From the fact that John is not dishonest we cannot conclude that John is honest, but only that he is
possibly so. ([26, p. 255])
The alternative is namely that John is neither dishonest, nor honest, so “John is not dishonest”, if true,
tells us what John fails to be and what the alternatives are: (i) being honest, (ii) neither being honest nor
being dishonest. The contradictory property is that it is not the case that it is not the case that John is
honest, which is logically equivalent to him being honest. More specifically, this quotation has inspired us to
introduce the trifurcation of cases presented in the Introduction. This trifurcation is epistemic rather than
ontological, as it bears on the (in-)validity of various inferences.
The definition of property negation must encapsulate the contrariety clause that the intensional negation
of one of two conjuncts that are mutually exclusive does not entail the truth of the other conjunct.
Definition 7 (Contrary properties). Let xι;F, G ()τω . Then the properties F,Gare mutually
contrary iff
wtx[[Fwt x]⊃ ¬[Gwt x]] ∧ ∃wtx[¬[Fwt x]∧ ¬[Gwt x]]
The definition states that it is not possible for xto co-instantiate Fand G, and possibly xinstantiates
neither F, nor G. The left-hand conjunct,
wtx[[Fwt x]⊃ ¬[Gwt x]]
is the clause that Fand Gare mutually exclusive. The second conjunct,
wtx[¬[Fwt x]∧ ¬[Gwt x]]
is the contrariety clause that the negation of one of the conjuncts [Fw t x], [Gwt x] does not entail the truth
of the other one.
Next, we want to show that any property [MpF] formed from a property Fby a modifier Mpprivative
with respect to Fis a property contrary to F. First, we prove the left-hand conjunct:
wtx[[[MpF]wt x]⊃ ¬[Fwt x]]
To this end, we apply the second clause of the definition of privative modifiers (Definition 6): p[[[0
Ess F ]p]
[[0
Ess [MpF]] λwλt [λx ¬[pwt x]]]]. Hence the property [MpF] has among its requisites at least one property
contradictory to a requisite of the property F. Let these properties be Pand λwλt [λx ¬[Pwt x]], respectively.
Then at no hw, tiis there an individual xthat would satisfy both [[MpF]wt x] and [Fwt x]; if there were such
an x, then according to Definitions 4 and 5, xwould also have to satisfy both [Pwt x] and ¬[Pwt x], which
is logically impossible.
Remark. This proves that the previous definition found in [6, §4.4] and [28] is a corollary of the new
Definition 6.
The contrariety clause wtx[¬[[MpF]wt x]∧ ¬[Fwt x]] holds due to the thesis of individual anti-
essentialism which we subscribe to: no individual has any purely contingent property necessarily.
We should not forget, however, the limiting case where Fis a trivial, non-contingent property with a
constant extension, such as being self-identical. In this case, necessarily, when the type is (say) ι,Fwt is the
entire type ιand λ x ¬[Fwt x] is an empty ι-set, because at no hw, tiis there an individual that would be
neither identical with itself nor non-identical with itself. Another example of a non-contingent property is
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.19(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 19
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
the property being identical to a or b.28 At all hw, tithe extension of this property is the set {a, b}, and
at no hw, tiis there an individual that would be neither identical with aor b, nor non-identical with aor
b, for both aand bare necessarily around to instantiate this property. The upshot is that non-contingent
properties do not lend themselves to being modified by privative modifiers on pain of necessary falsehood.
Hence, if Tri is such a trivial non-contingent property then the extension of [MpT ri] is necessarily the
empty ι-set for any modifier privative with respect to Tri. Such a modifier turns Tri into an ‘idle property’
that has necessarily an empty extension.
Definition 8 (General modifier privative with respect to a property f). Let =: (o(()τ ω ()τ ω )(()τ ω ()τ ω ))
be the identity relation defined over first-order modifiers, non (()τ ω ()τ ω ) a variable ranging over
first-order modifiers, f()τ ω ,Con : (o()τ ω ()τ ω ) the relation of contrariety between properties.
Then:
0
Non =λf λwλt [λx non [[[non f ]wt x][0
Con [non f ]f]]]
is the general modifier privative with respect to f.
Remark. Any of the modifiers non meeting the condition specified by Definition 8 are privative with respect
to the property F. Property negation takes a particular property Fto an arbitrary property contrary to it,
[non F ].29
Non is thus the unique general privative modifier, and it takes a property Fto the general contrary
property [0N on F ]. For instance, [0Non Banknote] is the general property contrary to the property of being
a banknote. Necessarily, the extension of [0N on Banknote]wt includes the extensions of the properties forged
banknote,banknote dissolved in acid,Monopoly banknote, etc., some of the extensions possibly being empty.
One might worry that it is too much to claim that, necessarily, the extension contains the full panoply of
non-banknotes. But it follows from Definition 8 that the full panoply is indeed involved. At any hw, ti, for
any individual xand the property constructed by F()τω , this holds:
[[0
N on F ]wt x] = non [[[non F ]wt x][0
Con [non F ]F]]
Hence, individual ahas the property [0
N on F ] iff ahas any property [non F ] for some non privative with
respect to F. Thus, the set [0
N on F ]wt is almost as large as the complement Fwt of the set Fwt. At some,
but not all, hw, tiit is the case that [0
N on F ]wt =Fwt . Or, when Fwt happens to be the entire type ι,
then [0
N on F ]wt must be the empty ι-set, i.e. the union of all empty sets [0
N on F ]wt. Definition 8 does not
exclude such modifier functions as do not even have a name in our vernacular.
Contrariety provides the weaker form of negation that is suitable for privative modifiers as explained
above. Definition 8 thus justifies the elimination rule Priv for modifiers Mpprivative with respect to property
Fstated above:
[[MpF]wt x][[0
N on F ]wt x]
The conclusion of Priv states that the predication of Feludes x:Fdoes not get to be predicated of x. For
instance, if the premise is that ais a fake banknote then the conclusion is that ais a Non-banknote, therefore
the property banknote is not predicated of a. Or if bis a wooden horse then bis a Non-horse. But if cis
28 TIL comes with a constant domain. See [6, 378–379].
29 Martin [24, p. 449] says, “Semantically [infinite negation, e.g. non-human] converts a term into one that stands for its non-empty
complement . . . ”. Our prop erty negation do es not come with an ontological restriction such as non-emptiness. However, more
importantly, our ‘non-F’ does not denote a complement set, but a contrary property; what denotes a set is ‘non-Fw t .
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.20(1-24)
20 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
awooden bird then it follows neither that cis a Non-horse, nor that cis a horse, because the properties
Non-horse and horse necessarily have still something in common (at least one common requisite), unlike
the properties wooden bird and horse or Non-horse.
Moreover, the partial order defined on sets of requisites makes it possible to compare how close the
privatively modified property [MpF] is to F. Since [MpF] and Fhave some requisites in common, they are
relatives. For instance, a fake banknote is not a fake passport or even a Non-passport; they are not relatives.
Yet a fake banknote is a close relative of banknotes, closer than, for instance, a Monopoly banknote or a
burnt banknote. From this point of view the most distant relative of the property Fis thus the property
[0
N on F ].
Let us run a test case. Can a paradox be deduced from our theory? Consider this example:
(1) Individual ais a 10 banknote
(2) Whatever is a 10 banknote is a banknote
(3) ais a banknote
(a) ais a forged 100 banknote created by adding a ‘0’ to ‘10’ to form ‘100’
(b) Whatever is a forged banknote is a non-banknote
(c) ais a non-banknote
Contradiction: (3) and (c).
This does not follow, however. One fact is that ais a tampered-with 10 banknote. Yet having a zero
add to ‘10’ does not have to undermine a’s property as a 10 banknote. Hence, amay remain a banknote,
for the modifier 10 is subsective with respect to the property of being a banknote. Another fact is that a
is a forged 100 banknote. From this, however, it does not follow that ais no longer a banknote. It only
follows that ais not a 100 banknote. The property that has been compromised by the attempted forgery
is that of being a 100 banknote, not the property of being a banknote per se. The apparent paradox
arises, because premise (b) fails to state that ais a forged 100 banknote and hence a non-100 banknote.
Therefore, premise (b) becomes irrelevant. Hence, acan be a 10 banknote (and thus a banknote) while
being a non-100 banknote.30
3.3. Double privation
We turn next to double privation, which has this form:
[M
p[MpF]]
Since Mpis privative with respect to [MpF], the intersection of the essences of [M
p[MpF]] and [MpF]
must be non-empty. And since Mpis privative with respect to F, the intersection of the essences of [MpF]
and Fmust also be non-empty. One may then wonder whether the respective essences of [M
p[MpF]] and F
can be disjunctive. We think not. There must be an overlap of requisites, and not just of any old properties,
but of carefully chosen ones.
Recall the earthquakes in central Italy in 2016. Many houses, bridges and other buildings and construc-
tions were damaged, some beyond repair. A demolished damaged house is surely not a house, but debris: a
particular object goes through the stages of being a house, then a damaged house and finally a demolished
damaged house, which is in material terms nothing but debris. Yet a demolished damaged house is different
from a demolished damaged bridge. A demolished damaged house shares requisites with houses that it does
not share with demolished damaged bridges.
30 We are indebted to Nikolaj Nottelmann and Lars Binderup for discussion of this example.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.21(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 21
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
It may so happen that the essence of [M
p[MpF]] is a superset of the essence of F. In such a case,
if xinstantiates [M
p[MpF]] then xalso instantiates F. For instance, a repaired damaged house is again
a house. To repair a damage is to undo the damage and in so doing returning the previously damaged
artefact to its still earlier state of functioning properly; such is the semantics of the verb ‘to repair’ and
the adjective ‘repaired’. So here we have come full circle back to F. This particular instance of the modifier
repaired is privative with respect to damaged house, because what is a non-house turns into a house. (We
are presupposing, to get the example off the ground, that a damaged house is so damaged that it no longer
qualifies as a house.) Being a repaired damaged house is one way of being a house. Formally:
[[0
Ess 0
House][0
Ess [0
Repaired [0
Damaged 0
House]]]]
Yet it may also so happen that the essence of [M
p[MpF]] and the essence of Fhave a non-empty
intersection, but neither is a subset of the other. For instance, a demolished damaged house is neither a
damaged house, nor a house, but something altogether different, namely a pile of rubble. The modifier
demolished, like repaired above, is privative with respect to damaged house, but the logical effect of applying
it to damaged house is the opposite. The semantics of the verb ‘to demolish’ puts it in opposition to ‘to repair’
or ‘to restore’. Nonetheless, a demolished damaged house must possess the requisite of having previously
been a house.
As is seen, the property of being a demolished damaged house spans three states: first, being a house;
second, being a damaged house; third, being a demolished damaged house. Formally:
[0
Ess [0
Demolished [0
Damaged 0
House]]] [0
Ess 0
House]6=
Absent the requisite property of having been previously a house, there is nothing to block the inference that
a demolished damaged house is (say) a demolished damaged bridge.
3.4. Three case studies
Here we revisit three examples that were broached above. For better readability of the following formulae,
we will now abbreviate formulae for constructions of the form ‘λwλt [λx ¬[pwt x]]’ as ‘not-p.
3.4.1. First fork
Since damaged is privative with respect to house, we have (as per Definition 6):
[[0
Ess 0
House][0
Ess [0
Damaged 0
House]]] 6=
∧ ∃p[[[0
Ess 0
House]p][[0
Ess [0
Damaged 0
House]] not-p]]
Hence damaged has turned some of the requisites of house into their opposites. For instance, if one of the
requisites of being a house is the property of being a place to live in, then damaged turns this property into
the property of not being a place to live in. Since repaired is privative with respect to the property damaged
house, we have:
[[0
Ess [0
Damaged 0
House]] [0
Ess [0
Repaired [0
Damaged 0
House]]]] 6=
∧ ∃q[[[0
Ess [0
Damaged 0
House]] q]
[[0
Ess [0
Repaired [0
Damaged 0
House]]] not-q]]
Now repaired cancels the effect of damaged; it must turn all those opposites not-p of damaged house back
into the original requisites pof House. Thus, among those properties qthat are contained in the essence
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.22(1-24)
22 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
of [0
Damaged 0
House] and appear as not-q in the essence of [0
Repaired [0
Damaged 0
House]] there must be
all those properties pwhich are contained in the essence of house and their opposites not-p in the essence
of [0
Damaged 0
House]. As a result, among these properties qthere are the properties λwλt [λx ¬¬[pwt x]],
hence p. We obtain:
[[0
Ess 0
House][0
Ess [0
Repaired [0
Damaged 0
House]]]]
The property repaired damaged house has all the requisites of house, being again a place to live in.
3.4.2. Second fork
Contrast the above property with the property demolished damaged house. Whatever is a demolished
damaged house cannot be a house, for the same reason that a demolished house cannot be a house. As soon
as we understand the meaning of the predicate ‘is a demolished damaged house’, we are able to calculate
which way it goes, and that we must land on the second fork. So, we know that a demolished damaged
house is a non-house. But we know something positive about it, too: we know that it is now a pile of rubble.
A demolished damaged house has been physically reduced to its raw matter (wood, steel, brick, etc.), just
like a melted-down statue is reduced to its raw matter (bronze, clay, etc.). The internal link between being
a demolished damaged house and being a pile of rubble is that that pile of rubble has a noble past as a
damaged house and before that as a house.
3.4.3. Third fork
Consider again former heir apparent. This combination of privatives is doubly dynamic due to the
backward-looking aspect of former and the forward-looking aspect of apparent, in the special sense of
‘apparent’ as ‘designated to become’. Someone who is a designated Fis currently not yet an F, though they
are supposed to become one. We are deploying the strict interpretation of ‘former’ as a privative rather
than a modal modifier to get the example of former heir apparent off the ground.31 With that in place,
someone who is a former heir apparent is not an heir, for one of two reasons: either the person succeeded
in succeeding the previous monarch (promotion), or the person is no longer being even considered for the
throne (demotion).
Accordingly, one requisite which heir apparent comes with is that any bearer must lack the property
of being the successor (where it is understood which is the relevant royal position, e.g. the office of King
of Denmark): this requisite property is due to the modification provided by apparent. Another requisite
which former heir apparent comes with is that any bearer must lack the property of being any longer the
prospective heir. The backward-looking aspect of former voids the forward-looking aspect of apparent, which
brings us to the present time where the bearer of the property of being a former heir apparent may, or may
not, be sitting on the throne.
3.4.4. Summary
To sum up these three case studies, which fork is the right one depends on the semantics of the modifiers
involved. When faced with iterated privatives, the agents who operate within some interactive system
for reasoning on the basis of natural-language texts can request additional information about particular
modifiers.32 The appropriate answer will be a refinement of the modifier in question. For instance, an
appropriate refinement of repaired would be this:
p[[[[0
Ess F ]p][[0
Ess[MpF]] not-p]] [[0
Ess [0
Repaired [MpF]]] p]]
31 On the privative reading, from “ais a former F” it can be inferred that ais no longer an F, hence is not an F. On the modal
reading, it cannot be excluded that ahas been reinstated as an F.
32 A particular such system is investigated in, e.g., [4] and [10].
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.23(1-24)
B. Jespersen et al. / Journal of Applied Logic ••• (••••)•••••• 23
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Thus, we can infer that whatever xis a repaired [MpF] is also an F. Similarly, a supplementary piece of
information about the semantics of demolished might be this:
p[[[[0
Ess F ]p][[0
Ess[MpF]] not-p]] [[0
Ess [0
Demolished [MpF]]] not-p]]
Then we can infer that a demolished [MpF] is not an F. If no such refinement can be supplied, then we
cannot decide which of the first two forks an individual xlands on, and so we know that we are facing a
case of the third fork.
4. Conclusion
The results obtained in this paper amount to an extension of the standard theory of property modification
by adding a logic of iterated privation to it. We started out with the problem that the received rule of single
privation is too crude, because it turns the root property into the contradictory property. To start solving
the problem, we replaced Boolean negation by property negation, enabling us to operate on contrary rather
than contradictory properties.
We then assigned so-called requisites to properties, and defined the essence of a property as the set of
all its requisites. Also, properties formed by means of iterated privation are equipped with requisites. They
underpin our presuppositional theory of positive predication, which is the predication of properties an object
must have, as a matter of analytic necessity, if it has a particular privatively modified property.
The notion of requisite properties enabled us to show that properties formed from iterated privatives give
rise to a trifurcation of cases between returning to the original root property or to a property contrary to
it or being semantically undecidable for want of further information. We have thereby exceeded the general
insight that pairs of privatives yield contraries rather than contradictories, because we are in a position to
calculate which of the forks of the trifurcation we land on.
Acknowledgements
This research has been supported by a University of Padua project on Disagreement, PRAT UNIPD
(Massimiliano Carrara), as well as by the Grant Agency of the Czech Republic Project No. GA15-13277S,
Hyperintensional Logic for Natural Language Analysis, and Project SGS No. SP2017/133 of the internal
grant agency of VŠB-TUO, Knowledge Model ling and its Applications in Software Engineering III (Marie
Duží and Bjørn Jespersen). Various versions of this material have been presented by Bjørn Jespersen as
invited lectures at the Heinrich Heine University Düsseldorf, 21 December 2017; at the Network of Danish
Philosophers Abroad, University of Southern Denmark, 8–9 September 2017; conference How To Say ‘Yes’ or
‘No’: Logical Approaches to Modes of Assertion and Denial, Universitá del Salento, Lecce (21–22 January
2016); Ruhr-Universität Bochum (5 November 2013); Hong Kong University (8 May 2012); Hong Kong
Polytechnic University (7 May 2012); as a tutorial at Technical University of Ostrava (26 February 2013);
and, as solicited lectures at LOGICA 2014, Hejnice (17–20 June 2014) together with Massimiliano Carrara;
CLPS 13, Ghent (16–18 September 2013). This research on iterated privation grew out of a project aimed at
developing an intensional logic for reasoning about technical artefacts in general and technical malfunction
in particular that Massimiliano Carrara and Bjørn Jespersen embarked upon in 2008 when the latter was
affiliated with the then-Section of Philosophy, Delft University of Technology. The Section sponsored a
one-month stay in Padua to initiate the project in cooperation with Massimiliano Carrara. Marie Duží later
joined the sub-project on iterated privation. The present paper completes what we would occasionally refer
to as ‘the malfunction trilogy’. We wish to thank two anonymous reviewers for Journal of Applied Logic for
very valuable comments and suggestions which improved the quality of the paper.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.24(1-24)
24 B. Jespersen et al. / Journal of Applied Logic ••• (••••)••••••
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
References
[1] K.R. Beesley, Evaluative adjectives as one-place predicates in Montague Grammar, J. Semant. 1 (1982) 195–249.
[2] M. Carrara, M. Soavi, Copies, replicas and counterfeits of artworks and artifacts, Monist 93 (2010) 417–435.
[3] S.S. Coulson, G. Fauconnier, Fake guns and stone lions: conceptual blending and privative adjectives, in: B. Fox, D.
Jurafsky, L. Michaelis (Eds.), Cognition and Function in Language, CSLI, Palo Alto, CA, 1999.
[4] M. Číhalová, M. Duží, N. Ciprich, M. Menšík, ‘Agents’ reasoning using TIL-Script and Prolog’, in: Information Modelling
and Knowledge Bases XXI , in: Frontiers in Artificial Intelligence and Applications, vol. 206, 2010, pp. 135–154.
[5] L. Del Frate, Failure: Analysis of an Engineering Concept, PhD Thesis, Delft University of Technology, Delft, 2014.
[6] M. Duží, B. Jespersen, P. Materna, Procedural Semantics for Hyperintensional Logic, Springer, Dordrecht, 2010.
[7] M. Duží, How to unify Russellian and Strawsonian definite descriptions, in: R. Ciuni, H. Wansing, C. Willkommen (Eds.),
Recent Trends in Philosophical Logic, vol. 41, 2014, pp. 85–101.
[8] M. Duží, Presuppositions and two kinds of negation, Log. Anal. 239 (2017) 245–263, special issue on How to Say ‘Yes’ or
‘No’, D. Chiffi (ed.).
[9] M. Duží, M. Číhalová, Questions, answers and presuppositions, Comput. Sist. 19 (2015) 647–659.
[10] M. Duží, M. Číhalová, M. Menšík, Communication in a multi-agent system; questions and answers, in: 13th International
Multidisciplinary Scientific GeoConference SGEM 2013 , Albena, Bulgaria, 2013, pp. 11–22.
[11] M. Enç, Towards a referential analysis of temporal expressions, Linguist. Philos. 9 (1986) 405–426.
[12] L. Floridi, N. Fresco, G. Primiero, On malfunctioning software, Synthese 192 (2015) 1199–1220.
[13] L.R. Horn, A Natural History of Negation, University of Chicago Press, London, Chicago, 1989.
[14] L.R. Horn, Duplex negatio affirmat? On the economy of double negation, in: Papers from the Parasession on Negation,
in: CSL, vol. 27, Chicago Linguistic Society, Chicago, 1991, pp. 80–106, pt. 2.
[15] L.R. Horn, Lie-toe-tease: double negatives and unexcluded middles, Philos. Stud. (2017), https://doi.org/10.1007/s11098-
015-0509-y.
[16] Ł. Iwańska, Reasoning with intensional negative adjectivals: semantics, pragmatics, and context, Comput. Intel l. 13 (1997)
348–390.
[17] B. Jespersen, Left subsectivity: how to infer that a round peg is round, Dialectica 70 (2016) 531–547.
[18] B. Jespersen, The phone booth puzzle, Organon F 13 (2006) 411–438.
[19] B. Jespersen, M. Carrara, A new logic of technical malfunction, Stud. Log. 101 (2013) 547–581.
[20] B. Jespersen, M. Carrara, Two conceptions of technical malfunction, Theoria 77 (2011) 117–138.
[21] B. Jespersen, P. Materna, Are wooden tables necessarily wooden?, Acta Anal. 17 (2002) 115–150.
[22] H. Kamp, Two theories about adjectives, in: E.L. Keenan (Ed.), Formal Semantics of Natural Language, Cambridge
University Press, Cambridge, 1975, pp. 123–155.
[23] D.C. Makinson, Topics in Modern Logic, Methuen and Co., London, 1973.
[24] J.N. Martin, All brutes are subhuman: Aristotle and Ockham on private [sic] negation, Synthese 134 (2003) 429–461.
[25] R. Montague, English as a formal language, in: B. Visentini (Ed.), Linguaggi nella societá e nella tecnica, Edizioni di
Communita, 1970, pp. 188–221.
[26] M. La Palme Reyes, et al., Models for non-Boolean negations in natural languages based on aspect analysis, in: D. Gabbay,
H. Wansing (Eds.), What is Negation? , Springer, Dordrecht, 1999, pp. 239–260.
[27] B. Partee, Privative adjectives: subsective plus coercion, in: R. Bäuerle, et al. (Eds.), Presuppositions and Discourse,
Elsevier, Amsterdam, 2001.
[28] G. Primiero, B. Jespersen, Alleged assassins: realist and constructivist semantics for modal modifiers, in: International Tbil-
isi Symposium on Logic, Language, and Computation, in: Lecture Notes in Computer Science, vol. 7758, 2013, pp. 94–114.
[29] M.E. Siegel, Capturing the Adjective, PhD Thesis, University of Massachusetts, Amherst, 1976.
[30] P. Tichý, Existence and god, J. Philos. 76 (1979) 403–420, reprinted in: V. Svoboda, B. Jespersen, C. Cheyne (Eds.),
Pavel Tichý’s Collected Papers in Logic and Philosophy, University of Otago Press, Filosofia, Czech Academy of Sciences,
Dunedin, Prague, 2004.
.
ARTICLE IN PRESS
UNCORRECTED PROOF
Please cite this article in press as: B. Jespersen et al., Iterated privation and positive predication, J. Appl. Log. (2017),
https://doi.org/10.1016/j.jal.2017.12.004
JID:JAL AID:484 /FLA [m3L; v1.226; Prn:9/12/2017; 10:11] P.25(1-24)
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 26
27 27
28 28
29 29
30 30
31 31
32 32
33 33
34 34
35 35
36 36
37 37
38 38
39 39
40 40
41 41
42 42
43 43
44 44
45 45
46 46
47 47
48 48
Sponsor names
Do not correct this page. Please mark corrections to sponsor names and grant numbers in the main text.
Grant Agency of the Czech Republic,country=Czech Republic, grants=GA15-13277S
VŠB-TUO,country=Czech Republic, grants=SP2017/133
... Duží and Fait [17] adjusted Gentzen's system of natural deduction for TIL so that the system can answer not only Yes-No questions by keywords searching but also Wh-questions by inferring computable knowledge from natural-language texts. Though the paper does not deal with the 24 For details on property modifiers, see Jespersen et al. [38] or Duží [13]. classification of Wh-questions, a useful logical technique of their answering is described here. ...
... 2022/(oτ ); ≤ τ /(oτ (oτ )): ≤ τ stands for the relation between the evaluation time t and time interval of the year 2022 such that t comes before the end of the year 2022. 38 The path with the statement 'else fail' means that the denoted proposition evaluates to no truth value. ...
... p w → (oτ )). The class of those intervals d that have a non-empty intersection with a reference 38 More on dealing with time and calendars can be found in [23]. 39 A detailed analysis of particular kinds of tenses can be found in [19] (Section 2.5.2). ...
Article
In a multiagent and multi-cultural world, the fine-grained analysis of agents’ dynamic behaviour, i.e. of their activities, is essential. Dynamic activities are actions that are characterized by an agent who executes the action and by other participants of the action. Wh-questions on the participants of the actions pose a difficult particular challenge because the variability of the types of possible answers to such questions is huge. To deal with the problem, we propose the analysis and classification of Wh-questions apt for agents’ communication in a multiagent system (MAS). Our proposal of such a system consists of agents who communicate with their fellow agents by messaging so that each autonomous agent, though resource-bounded, can make less or more rational decisions to meet its own and collective goals. In addition, by communicating with other fellow agents and their environment, agents can learn new concepts and enrich their ontology so that their behaviour is dynamic. We aim to make a general proposal of the system so that the ‘envelope’ of agents’ messages can be formalized in any MAS standard, be it The Foundation for Intelligent Physical Agents - Agent Communication Language (FIPA-ACL) or Knowledge Query and Manipulation Language (KQML). Yet, the content of messages is encoded in a formalized natural language. To this end, we apply Transparent Intensional Logic (TIL) with its procedural semantics which is particularly apt for a fine-grained analysis in which all the semantically salient features of natural language can be plausibly formalized. In this paper, we concentrate on analysing the content of query messages, particularly the content of those that encode Wh-questions and the answers to them. We also summarize TIL deduction system that makes it possible to answer such questions in an intelligent way. Linguists distinguish several subtypes of Wh-questions. Though the linguistic classification is helpful, it is not always suitable for agents’ communication and reasoning. We need the classification based on a logical analysis of Wh-questions so that the agents can infer possible answers to such questions rather than only looking for them by keywords. This paper aims to apply an appropriate classification of the logical types of Wh-questions and the analysis of such questions; we concentrate in particular on questions concerning the participants of activities. The application of these results to the analysis of processes and events based on verb valency frames is another novelty of the paper.
... Another notion we need to analyze due to our example of the fake banknote that is a banknote is that of property modifier. 17 This example involves a privative modifier, fake. TIL defines modifiers by way of requisites. ...
... Various versions of relevant / relevance logic have been put forward to address this sort of issue. See, for instance,[4].17 See[6] for the TIL treatment of modifiers and[17] for privative modifiers in particular. ...
... See, for instance,[4].17 See[6] for the TIL treatment of modifiers and[17] for privative modifiers in particular. ...
Article
We talk about 'impossible objects' in many areas, ranging from empirical and non-empirical theories to the realm of fiction, myth and folklore: a mathematical pendulum, a perfect market, the set of all sets that are not members of themselves, Kafka's Gregor Samsa, Pegasus, The Puss in the Boots, and so forth. This paper proposes a hyperintensional account of a special case of impossible objects, so-called 'impossible individuals'. Our (broadly Fregean) proposal is to identify 'impossi-ble individuals' with necessarily empty individual concepts. The main goal of the paper is to develop a method that enables us to discover inconsistencies in specifications of individual concepts and thus prove that such concepts could not possibly be matched by an extension (an individual). Furthermore, this approach allows for a fine-grained individuation of impossible individuals. Fine-graining will not be explored in the present paper, but its very possibility adds to the overall plausibility of the present account.
... Thus, 'the Pope' and 'Bishop of Rome' are not synonymous terms and cannot be mutually substituted here, because in a hyperintensional contexts only synonymous terms with procedurally isomorphic meanings can be mutually substituted. 16 16 The relation of procedural isomorphism has been introduced in TIL to deal with the problem of the structural isomorphism of meanings, hence of cohyperintensionality, hence of synonymy. It has been demonstrated that the individuation of procedures assigned to expressions as their structured meaning cannot be decided in virtue of a universal criterion applicable to Hence, existential quantifying into hyperintensional contexts is valid only if we quantify over objects presented by Trivialization. ...
... For details on property modifiers see, for instance,[5] and[16]. ...
Article
In this paper, we introduce the system for inferring implicit computable knowledge from textual data by natural deduction. Our background system is Transparent Intensional Logic (TIL) with its procedural semantics that assigns abstract procedures known as TIL constructions to terms of natural language as their context-invariant meanings. The input data for our method are produced by the so-called Normal Translation Algorithm (NTA). The algorithm processes natural-language texts and produces TIL constructions. In this way we have obtained a large corpus of TIL meaning procedures. These procedures are furthermore processed by our algorithms for type checking and context recognition, so that the rules of natural deduction for inferring computable knowledge can be afterwards applied.
... In the case of iterated privation, privative modifiers are replaced by the general privative modifier Non. See (Jespersen et al. 2017). ...
... 14 For details on property modifiers see, for instance, [5] and [16]. ...
... TIL analysis has been introduced in numerous papers, see, e.g. (Jespersen, Carrara, Duží, 2017), (Duží, 2017) or (Jespersen, 2015), (Jespersen, 2016). The issue we deal with bellow is the rule of left subsectivity. ...
... A detailed introduction to the TIL system is out of the scope of this paper. Referring to numerous papers and two books (see, e.g., [2], [3], [12], [15]), we just briefly summarize. ...
Chapter
Full-text available
In this paper, we deal with Leibniz’s rule of substitution of identicals, and describe how the rule can be applied in the TIL-Script language. The main goal is to introduce the algorithm of valid application of the substitution rules in all the three kinds of context that we distinguish in the TIL-Script language. The language is a computational variant of TIL, which is a hyperintensional, partial typed Open image in new window -calculus. Hyperintensional, because the meaning of TIL Open image in new window -terms are procedures producing functions rather than the denoted functions themselves. Partial, because TIL is a logic of partial functions, and typed, because all the entities of TIL ontology receive a type. Based on the results of context recognition the algorithm makes it possible to validly apply the substitution rules and derive relevant new pieces of analytic information.
Chapter
Full-text available
The paper deals with natural language processing and question answering over large corpora of formalised natural language texts. Our background theory is the system of Transparent Intensional Logic (TIL) which is a partial, hyperintensional, typed λ-calculus. Having a fine-grained analysis of natural language sentences in the form of TIL constructions, we apply Gentzen’s system of natural deduction adjusted for TIL to answer questions in an ‘intelligent’ way. It means that our system derives logical consequences entailed by the input sentences rather than merely searching answers by keywords. The theory of question answering must involve special rules rooted in the rich semantics of a natural language, and the TIL system makes it possible to formalise all the semantically salient features of natural languages in a fine-grained way. In particular, since TIL is a logic of partial functions, it is apt for dealing with non-referring terms and sentences with truth-value gaps. It is important because sentences often come attached with a presupposition that must be true so that a given sentence had any truth-value. And since answering is no less important than raising questions, we also propose a method of adequate unambiguous answering questions with presuppositions. In case the presupposition of a question is not true (because either false or ‘gappy’), there is no unambiguous direct answer, and an adequate complete answer is instead a negated presupposition. There are two novelties; one is the analysis and answering of Wh-questions that transform into λ-terms referring to α-objects where α is not the type of a truth-value. The second is integration of special rules rooted in the semantics of natural language into Gentzen’s system of natural deduction, together with a heuristic method of searching relevant sentences in the labyrinth of input text data that is driven by constituents of a given question.
Chapter
We talk about ‘impossible objects’ in many areas, ranging from empirical and non-empirical theories to the realm of fiction, myth and folklore: a mathematical pendulum, a perfect market, the set of all sets that are not members of themselves, Kafka’s Gregor Samsa, Pegasus, The Puss in the Boots, and so forth. This paper proposes a hyperintensional account of a special case of impossible objects, so-called ‘impossible individuals’. Our (broadly Fregean) proposal is to identify ‘impossible individuals’ with necessarily empty individual concepts. The main goal of the paper is to develop a method that enables us to discover inconsistencies in specifications of individual concepts and thus prove that such concepts could not possibly be matched by an extension (an individual). Furthermore, this approach allows for a fine-grained individuation of impossible individuals. Fine-graining will not be explored in the present paper, but its very possibility adds to the overall plausibility of the present account.
Chapter
The existing security attack and defense mechanisms are based on positive logic system and there are some disadvantages. In order to solve the disadvantages of the existing mechanisms, the new security attack and defense mechanisms based on negative logic system and its applications are innovatively proposed in this paper. Specifically speaking, at first, we propose the negative logic system which is totally new to the security area. Then, we propose the security attack and defense mechanisms based on negative logic system and analyze its performance. Moreover, we introduce the specific applications of attack and defense mechanisms based on negative logic system and take the active probe response processing method and system based on NLS for detailed description. With the method and new security attack and defense mechanisms based on NLS in this paper, its advantages are as follows. It can improve security from the essence of cyber attack and defense and have great application value for security. It can be applied in active probe response processing area, secret sharing area, etc. Most importantly, it can improve security of all these areas, which is of great significance to cyberspace security.
Article
Full-text available
In this paper I deal with sentences that come with a presupposition that is entailed by the positive as well as negated form of a given sentence. However, there are two kinds of negation, namely narrow-scope and wide-scope negation. I am going to prove that while the former is presupposition-preserving, the latter is presupposition-denying. Thus the main contribution of this paper is the proof that these two kinds of negation are not equivalent. This issue has much in common with the difference between topic and focus articulation within a sentence. Whereas articulating the topic of a sentence activates a presupposition, articulating the focus frequently yields merely an entailment. My background theory is Transparent Intensional Logic (TIL). TIL is an expressive logic apt for the analysis of sentences with presuppositions, because in TIL we work with partial functions, in particular with propositions with truth-value gaps. Moreover, procedural semantics of TIL makes it possible to define a general analytic schema of sentences associated with presuppositions, which is another novel contribution of this paper.
Chapter
Full-text available
In this paper I will deal with ambiguities in natural language exemplifying the difference between topic and focus articulation within a sentence. I will show that whereas articulating the topic of a sentence activates a presupposition, articulating the focus frequently yields merely an entailment. Based on analysis of topic-focus articulation, I propose a solution to the almost hundred-year old dispute over Strawsonian versus Russellian definite descriptions. The point of departure is that sentences of the form ‘The \(F\) is a \(G\)’ are ambiguous. Their ambiguity stems from different topic-focus articulations of such sentences. Russell and Strawson took themselves to be at loggerheads, whereas, in fact, they spoke at cross purposes. My novel contribution advances the research into definite descriptions by pointing out how progress has been hampered by a false dilemma and how to move beyond that dilemma. The point is this. If ‘the \(F\)’ is the topic phrase then this description occurs with de re supposition and Strawson’s analysis appears to be what is wanted. On this reading the sentence presupposes the existence of the descriptum of ‘the \(F\)’. The other option is ‘\(G\)’ occurring as topic and ‘the \(F\)’ as focus. This reading corresponds to Donnellan’s attributive use of ‘the \(F\)’ and the description occurs with de dicto supposition. On this reading the Russellian analysis gets the truth-conditions of the sentence right. The existence of a unique \(F\) is merely entailed. This paper demonstrates how to unify these disparate insights into one coherent theory of definite descriptions.
Article
Full-text available
Litotes, “a figure of speech in which an affirmative is expressed by the negative of the contrary” (OED) has had some tough reviews. For Pope and Swift (“Scriblerus” 1727), litotes—stock examples include “no mean feat”, “no small problem”, and “not bad at all”—is “the peculiar talent of Ladies, Whisperers, and Backbiters”; for Orwell (1946), it is a means to affect “an appearance of profundity” that we can deport from English “by memorizing this sentence: A not unblack dog was chasing a not unsmall rabbit across a not ungreen field.” But such ridicule is not without equivocation, given that litotes, or “logical” (non-concordial) double negation, may or may not be semantically redundant. When the negation of a logical contrary yields an unexcluded middle, it contributes to expressive power: someone who is not unhappy may not be happy either, and an occurrence may not be infrequent without being frequent. But if something is not possible, what can it be but possible? Why does Crashaw’s “not impossible she” survive rhetorically while Orwell’s “not unsmall rabbit” is doomed? How is Robbie being “not not friends” with Mary on 7th Heaven distinct from being friends with her, if not not-p reduces to p? The key is recognizing in litotes a corollary of MaxContrary, the tendency for contradictory (wide-scope) sentential negation ¬p to strengthen (at least) pragmatically to a contrary ©p, as when the formal contradictory Fr. “Il ne faut pas partir” (lit. ‘It is not necessary to leave’) is reinterpreted as expressing a contrary (‘one must not-leave’). Just as the Law of Excluded Middle can apply where it “shouldn’t”, resulting in pragmatically presupposed disjunctions between semantic contraries, so that “p v ©p” amounts to an instance of “p v ¬p”, the Law of Double Negation can fail to apply where it “should”. When not not-p conveys ¬©p, the negation of a virtual contrary, the middle between p and not-p is no longer excluded, rendering the Fregean dictum that “Wrapping up a thought in double negation does not alter its truth value” not unproblematic.
Article
Full-text available
The paper deals with the analysis of empirical questions that come attached with a presupposition. Our thesis is this. In case that the presupposition of a question is not true, there is no unambiguous direct answer, and an adequate complete answer is a negated presupposition. Yet this simple idea is connected with a bunch of problems. First, we must distinguish between a pragmatic and semantic presupposition, and between a presupposition and mere entailment. Second, we show that the common definition of a presupposition of a question as such a proposition that is entailed by every possible answer to the question is not precise. We follow Frege and Strawson in treating survival under negation as the most important test for presupposition. But a negative answer to a question is often ambiguous. The ambiguity consists in not distinguishing between two kinds of negative answers, to wit the answers applying narrow-scope or wide-scope negation. While the former preserves presupposition, the latter seems to be presupposition denying. We show that in order that the negative answer be unambiguous, instead of the wide-scope negation presumably denying presupposition, an adequate and unambiguous answer is just the negated presupposition. Having defined presupposition of a question more precisely, we then examine Yes-No questions, Wh-questions and exclusive-or questions with respect to several kinds of presupposition triggers. These include inter alia topic-focus articulation, verbs expressing termination of an activity, factive verbs, the „whys and how comes“, and past or future tense with reference time interval. Our background theory is Transparent Intensional Logic (TIL) with its procedural semantics. TIL is an expressive logic apt for analysis of questions and presuppositions, because within TIL we work with partial functions, in particular with propositions with truth-value gaps. These features enabled us to define a general analytic schema of sentences associated with a presupposition. Our results are applicable in linguistics and artificial intelligence, in particular in the systems the behaviour of which is controlled by communication and reasoning of intelligent social agents.
Article
A property modifier is a function that takes a property to a property. For instance, the modifier short takes the property being a Dutchman to the property being a short Dutchman. Assume that being a round peg is a property obtained by means of modification, round being the modifier and being a peg the input property. Then how are we to infer that a round peg is a peg? By means of a rule of right subsectivity. How are we to infer that a round peg is round? By means of a rule of left subsectivity. This paper puts forward two rules (one general, the other special) of left subsectivity. The rules fill a gap in the prevalent theory of property modification. The paper also explains why the rules are philosophically relevant. © 2017 The Author dialectica
Article
In a 1997 paper Jennifer Saul adduces various examples of simple sentences in which the substitution of one co-referential singular term for another appears to be invalid. I address the question of whether anti-substitution is logically justified by examining the validity and soundness of substitution of co-referential singular terms in three simple-sentence arguments each exhibiting a different logical structure. The result is twofold. First, all three arguments are valid, provided Leibniz's Law is valid with respect to simple sentences (something Saul herself does not doubt). Thus, as far as these arguments are concerned, there is no logical problem with substitution in simple sentences. Second, two of the arguments cannot be sound, because their respective sets of premises are inconsistent. Thus, it would be logically irrational to commit oneself to all the premises of the respective arguments. To the extent that the origin of Saul's puzzles is in logic (rather than pragmatics, say), I suggest, tentatively, that substitution may appear to be invalid because the issues of validity and soundness have not been kept separate. I then consider in depth Saul's first sentence, "Clark Kent enters a phone booth and Superman exits". Obviously, two-way substitution is trivially valid, if the expressions are co-referential semantically (and not just grammatically) proper names, the conclusion being but a rephrasing of the premise. However, I argue that a non-trivial semantic analysis of this sentence should take account of the diachronicity of Clark Kent's entrance and Superman's exit while preserving the internal link between being Superman and being Clark Kent. I propose the following. 'Superman' and 'Clark Kent' refer to two distinct individual concepts. "Superman is Clark Kent" then no longer expresses the self-identity of an individual bearing two names, but that two named concepts are held together by the requisite relation: wherever and whenever someone falls under the concept of Superman the same individual also falls under the Clark Kent concept, whereas there are exceptions to the converse. This semantic analysis always validates the substitution of 'Clark Kent' for 'Superman', but validates the substitution of 'Superman' for 'Clark Kent' only if the additional condition is met that somebody should fall under the Superman concept when Clark Kent enters. The analysis is accompanied by a device of extensionalisation from individual concepts to individuals and two rules of predication.